[Linux-cluster] virtual address went down?

[Lon Hohberger]

On Wed, 2006-10-18 at 16:32 -0400, jason at monsterjam.org wrote:
> > > so Ive had a test cluster running for quite a while now, both nodes of a 2 node cluster are up, 
> > > but the virtual address seems to have disappeared.. its not pingable, neither server has it 
> > > configured anymore..
> > 
> > You're using 'ip addr list', not 'ifconfig' to look for it, right?
> actually, I was using ifconfig, I forgot that you need to use ip addr 
> list.. my mistake.. but the address was for sure not pingable.
> 
> > 
> > > The only application I had using the virtual address was apache (just for 
> > > testing it). what logs/information should I be looking at to see what happened and why?
> > 
> > Firewall bits?
> not sure what you mean here.. 

IPTables.  Though, if it was working and it's available via ip addr
list, it should still be working.

When this happens, does the IP still appear in 'ip addr list' ?  What
about running 'clustat' -- is the service started, or did it perhaps
fail?

> > What kind of ethernet cards do you have, out of curiosity, and are they
> > in any sort of bonding configuration?
> > 
> [jason at tf1 ~]$ lspci | grep -i ether
> 02:0c.0 Ethernet controller: Intel Corporation 82557/8/9 [Ethernet Pro 
> 100] (rev 0d)
> 06:07.0 Ethernet controller: Intel Corporation 82541GI/PI Gigabit 
> Ethernet Controller (rev 05)
> 07:08.0 Ethernet controller: Intel Corporation 82541GI/PI Gigabit 
> Ethernet Controller (rev 05)
> 
> im not doing any bonding.. 

Ok, there was/is a bug with e1000 in bonding configurations if you have
4+ of the cards in the system.  That's what I was looking for.  Since
your configuration has none of the above criteria, it's unlikely to be
that problem.

-- Lon